/home/karen/Documents/GitHub/Muscle_wasting

Analyzing the relevance of the mirnas¶

Hello Kasia and Pilib, here is my report of the microRNAs

Now we saw with the filtered and whole network Let's get the mirnas from the relevant PR score and then evaluate them on the whole network

Background¶

There are 3 main steps relevant for this report

  • DE data
  • Network (Transcripor factors)
  • Pathways

DE data¶

I joined the different rnaseq expression data and using dseq2 got the differentially expressed data. Is important to mention that the DE genes are different than the ones obteined doing the experiments individually (really small overlap). We had the comparisons - Young vs Old - Young vs Middle Age - Middle Age vs Old Therefore is always younger vs older, therefore, up regulared means that it decreases with age, and downregulated that increases with age.

Network¶

Using the DE genes, was possible to identify the enriched TF using collectri database and the decoupler library function run_ulm. Using the same database, is possible to build the full network of gene-TF interactions. Additionally, using my network, I added the microRNAs as well.

Pathways¶

Using msigdb we assigned to each gene the pathways they are involved. We only considered the pathays that were at least slightly enriched. From all the pathways, on this notebook, only the pathways that seem logically involved on sarcopenia are presented:

- ATP 	
- MITOCHONDRI 	
- RESPIRAT 	
- METABOLI 	
- OXIDATIVE_PHOSPHORYLATION 	
- NONALCOHOLIC_FATTY_LIVER 	
- MUSCLE 	
- ELECTRON

Now, as you can see, the name of the pathways are not compleate, this is because as microcondri I want to include terms as

- GOBP_MITOCHONDRIAL_ELECTRON_TRANSPORT_NADH_TO_UBIQUINONE
- GOCC_INNER_MITOCHONDRIAL_MEMBRANE_PROTEIN_COMPLEX
- HP_ABNORMAL_MITOCHONDRIA_IN_MUSCLE_TISSUE
- etc

All the pathways were added on the network with my software

Now that we have the data about the network, let's see the impact of the microRNAs

We used 2 netwoks here, the original one of 4159 nodes and 14587 edges with all the genes, TF and microRNAs generated, and the second one that is the same network but only the nodes whitin the top 80% of nodes of the pageRank rank. This limit was done given the fact that with that small theshold, from 566 mirnas to 13 mirnas.

MicroRNAs influence on genes¶

Starting on each of those 13 mirnas, we cover all the accesible paths from that node (each edge visited only once), if the influence is inhibition, it applies a -1, if it is activation, +1. Doing that, we can register in each time the mirna visitis a particular gene, see if the mirna has a positive or negative interaction. Each node has a list of (1,-1), the length of the list means the amount of times the mirna is able to reach the node. For each mirna, you can see the impact on each gene and looks like this:

MYC TERT SPI1 BGLAP IL2 SMAD7 E2F3
hsa-miR-21-5p [-1, -1, -1, 1, 1, 1, 1, -1, 1, 1, 1, 1, -1, 1... [-1, -1, -1, -1, -1, 1, 1, -1, 1, -1, -1, 1, 1... [-1, -1, 1] [1, 1, 1, -1, -1, 1, -1, -1] [-1, -1, -1, -1, -1, 1, -1, -1, 1, 1, -1, -1, ... [1, 1, 1, 1, 1, 1, -1] [1, 1, -1, 1, 1, 1, -1]
hsa-miR-210-3p [-1, -1, -1, -1, 1, 1, 1, -1, 1, -1, 1, -1, 1,... [-1, -1, 1, 1, -1, -1, -1, 1, 1, -1, 1, -1, 1,... [-1, -1, 1] [-1, 1, 1, -1, 1, 1, -1, -1] [-1, -1, -1, -1, -1, -1, 1, -1, -1, 1, 1, -1, ... [1, 1, 1, 1, 1, 1, -1] [-1, -1, -1, -1, 1, 1, -1]
hsa-miR-145-5p [-1, -1, -1, -1, -1, 1, -1, 1, -1, -1, -1, -1,... [-1, 1, 1, -1, 1, 1, 1, -1, -1, -1, -1, -1, -1... [-1, -1, -1] [1, 1, -1, -1, 1, 1, -1, -1] [-1, 1, 1, 1, 1, 1, 1, 1, -1, -1, -1, -1, -1, 1] [-1, -1, -1, -1, -1, -1, -1] [1, -1, 1, -1, 1, -1, 1]
hsa-miR-143-3p [-1, -1, -1, -1, -1, 1, 1, -1, 1, -1, -1, -1, ... [-1, -1, 1, 1, -1, 1, 1, 1, -1, -1, -1, -1, -1... [-1, -1, -1] [1, 1, -1, -1, 1, 1, -1, -1] [-1, 1, 1, 1, 1, 1, 1, 1, -1, -1, -1, -1, -1, 1] [-1, -1, -1, -1, -1, -1, -1] [1, 1, -1, 1, -1, 1, -1]
hsa-miR-93-5p
hsa-miR-375
hsa-miR-224-5p

interesting, hsa-miR-21-5p and hsa-miR-210-3p behaive really similar but is not the same, and hsa-miR-145-5p and hsa-miR-143-3p are similar too. Interesting is that (hsa-miR-21-5p and hsa-miR-210-3p) and (hsa-miR-145-5p and hsa-miR-143-3p) behaive almost opposite!!!

We can then look the data as a dotplot, the size of the dot means the amount of different ways the mirna has to reach that gene, if it is red, is that that mirna inhibits, if it blue, it means that the mirna activates it.

If the circle is big, but the color is white-ish, that means that some paths ibhibits and other activate. This is the dotplot of the 12895 genes... yeah, I recommend just scroll until the next section. (seriously, do cntl + f "next section")

12895
No description has been provided for this image

This is the "next section"!

Up or down reculated¶

Young vs Old¶

See the influcne of the genes that are down regualted in Young respect to old, this means that they INCREASE WITH AGE. Therefore, what we would like to see is more big red dots that means that the microRNA inhibits that gene.

Get the up or down reluation

This are only the 1381 genes that are down regulated and then the 589 genes that are up regulated

1381
No description has been provided for this image

See the influcne of the genes that are up regualted in Young respect to old, this means that they DECREASED WITH AGE. Therefore, what we would like to see is more big blue dots that means that the microRNA estimulate that gene.

589
No description has been provided for this image

Young vs Middle age¶

Down regulated in young respect of middle age

96
No description has been provided for this image

Up regulated in young respect of middle age

212
No description has been provided for this image

Middle age vs Old¶

These are the genes down regulated in middle age vs Old, this is the genes that increase with age.

544
No description has been provided for this image

These are the genes up regulated in middle age vs Old, this is the genes that decrease with age... oh, there were not.

Pathways affected¶

Now, usig those 13 microRNA nodes in the whole network, we start on that mirna and do on 5 steps, 10 paths, doing that we registered the paths that the path visits (I am still working on how to describe this with words) But basically we have record of all the pathways genes alterated.

We recoreded the times it gets on a gene on that pathway. There are registered on the table bellow. Since there are many pathways that we don't consider, there is a column called Different_pathways that shows the unique pathways the mirna affected somehow on those 5 steps, Total is the sum of all (considerered or not) and participation is the sum of those that are considered.

ATP MITOCHONDRI RESPIRAT METABOLI OXIDATIVE_PHOSPHORYLATION NONALCOHOLIC_FATTY_LIVER MUSCLE ELECTRON Different_pathways Total participation
hsa-miR-223 0 0 0 0 0 0 0 0 0 0 0
hsa-miR-21-5p 1 0 2 6 1 3 2 0 20 240 15
hsa-miR-210-3p 0 0 0 0 0 0 0 0 2 24 0
hsa-miR-122 0 0 0 0 0 0 0 0 0 0 0
hsa-miR-145-5p 0 1 0 2 0 1 1 0 12 152 5
hsa-miR-224-5p 0 0 0 0 0 1 0 0 2 16 1
hsa-miR-93-5p 0 1 0 0 0 0 0 0 2 16 1
hsa-miR-342 0 0 0 0 0 0 0 0 0 0 0
hsa-miR-221 0 0 0 0 0 0 0 0 0 0 0
hsa-miR-375 0 0 0 0 0 0 0 0 0 0 0
hsa-miR-17 0 0 0 0 0 0 0 0 0 0 0
hsa-miR-152 0 0 0 0 0 0 0 0 0 0 0
hsa-miR-143-3p 1 2 0 3 0 1 3 0 11 112 10

And here is the same data but on a heatmap. The most... brown(?) is the more times it land on that pathway.

<Axes: >
No description has been provided for this image

Mirnas that affect differential expressed genes¶

Some micrornas, on that 5 step path, affect DE genes on Young vs Old, Young vs Middle Age and Middle age vs Old.

Here are the mirnas que affect DE genes and how many of those genes it affects.

m_l m_s yo ym mo
hsa-miR-21-5p 0 0 9 2 3
hsa-miR-210-3p 0 0 3 0 1
hsa-miR-145-5p 0 0 10 0 2
hsa-miR-224-5p 0 0 1 0 0
hsa-miR-93-5p 0 0 2 0 0
hsa-miR-375 0 0 1 0 0
hsa-miR-143-3p 0 0 6 2 1